I love it. So exciting :D Looks like it’s almost entirely made out of printed parts.. just the motors and the screws/bolts left to go.

@stunmonkey: Actually, printed parts are by far the cheapest thing involved in a reprap. They max out at about $10/lb for materials and power – the current design has around three pounds at full infill. The sarrus-based design probably saves around $80 in bearings and rod alone, and all parts only really have one obvious way to go together for assembling the machine. We’ve yet to see whether it can achieve the requisite positioning accuracy, but it’s not attached to mainline anyways.

The “restriction” has more to do with a 3D printer being the cheapest machine to get and run available to produce the kind of parts it needs without needing skilled labor. Not using it, you either need a much more skilled operator to make your parts, or a much more expensive machine.

@stunmonkey

You may be interested in a series of books written by the late Dave Gingery, and continued by his son Vince. He gave complete, illustrated instructions on casting and fabricating all the major machine tools from scrap aluminium. This included the metal lathe, shaper, horizontal mill, drill press, accessories and upgrades. Additional books detailed a sheet metal slip-roller and brake, a bandsaw, and various other devices – like a banjo. Dave was pretty eclectic. His books are still carried by Lindsay Technical Books, and are a pretty good read. Lots of people have built his machines and have made improvements in the design and methods, but they are still fun & informative.

Looks cool, next step hexapod ;)

It reminds me of the “walking tanks” on Raiden II

@rallen

Dave Gingery had a really good set of concepts, and a good series of books.

Showing how to build an entire machine ship from nothing was brilliant, and it statred by making the simplest and most useful tool first – a lathe.

A lathe is far more practical for third world countries than is a 3D printer, is easier and cheaper to produce and use, and has the side benefit of being able to produce the parts for a 3D printer, as well.

Besides, if you can’t source the parts for even that locally, then how in the hell do you get clean power, a computer, software, computer-trained personnel, and the drive electronics as the precursor to the 3D printer?
The structure itself is almost meaningless logistically.

As well, a 3D printer is a great EXTENSION to a basic machine shop, not a replacement. The machine shop can work very well without a printer, but the printer is nigh-useless without a machine shop.

@ Everyone who is a RepRap fan

There is one aspect of this project that isn’t obvious, but is just as revolutionary a hack as everything else – the motor controller.

fdavies did not make any modifications to the standard RepRap controller software or skeinforge. Mr. Davies is using an arduino to control the DC motors and read the optoendcoders locally. The control signal being sent to that board is the standard stepper controller signals that come off the motor board as with any other RR-based printer. The control pulses come in, a step-count is made, and then the closed-loop control system moves the DC motors the correct amount. As a nice side-effect, you can take the gantry and yank it totally out of position, and it ziiiiip back into place immediately.

As far as efficiency goes, it’s a total loss. Adding another processor to the chain to avoid having to re-do firmware is terrible. However, being able to throw your own hardware under the standard RepRap software with a hardware-based hardware abstraction layer is pretty cool. I am hoping that this project will inspire/disgust some of the folks involved in the firmware design end of things to actually move to an abstracted-hardware design that is easier to modify at the end-user level.

Looks like this could be made with laser-cut 2D parts. I don’t think this arrangement is a true Sarrus Link. The Sarrus Link converts a limited circular motion to a linear motion without reference guide-ways (http://en.wikipedia.org/wiki/Sarrus_linkage). A true Sarrus link would be driven by a circular arc imposed on the apex of the bending joint. But a true Sarrus link would introduce non-linear mathematical complexity to the print engine code – and if you’re not careful varying precision at different points on in the work-space. The solution here doesn’t appear to have this issue. This solution does indeed remove the need for precision linear bearings and rods. Nicely-done.

@MattD

Remember: Given sufficient thrust, pigs fly quite well.

I recognise those motors!

I think they are the carriage drive motors from HP 6##, 7## or 8## inkjet printers. If you want some good but cheap DC motors, I can highly recommend getting them out of junk HP inkjet printers (I get mine from my local e-waste scrap dealer – very very cheap, and I give them back the printer to them to dispose of).

I am using the motors at about 20V for the drive motors for a robot (Asus wifi router with USB2.0 and OpenWRT Linux installed running Python pyserial and pyduino, Arduino and USB stick connected using USB hub, Arduino running Firmata).

As pointed out above, you also get precision bearing rods and very high precision optical encoders (linear or rotational depending on model).

I like it. This probably won’t become the mainstream RepRap, but this is a great way to expand the hacker workshop vocabulary, alongside your Gingery set of machines that you enhanced with CNC functionality from RepRap electronics. Also, if this is capable of printing the parts for a standard RepRap (I doubt it’s not), this might be a good RepStrap to keep around as a backup for when your main machine breaks something.

Also, I like cheap.